Studies of cultured cells have indicated that the mammalian target of rapamycin complex 1 (mTORC1) mediates the development of insulin resistance. Because a role for mTORC1 in the development of skeletal muscle insulin resistance has not been established, we studied mTORC1 activity in skeletal muscles of ob/ob (OB) mice and wild-type (WT) mice. In vivo insulin action was assessed in muscles of mice 15 min following an intraperitoneal injection of insulin or an equivalent volume of saline. In the basal state, the phosphorylation of S6K on Thr389, mTOR on Ser2448, and PRAS40 on Thr246 were increased significantly in muscles from OB mice compared with WT mice. The increase in basal mTORC1 signaling was associated with an increase in basal PKB phosphorylation on Thr308 and Ser473. In the insulin-stimulated state, no differences existed in the phosphorylation of S6K on Thr389, but PKB phosphorylation on Thr308 and Ser473 was significantly reduced in muscles of OB compared with WT mice. Despite elevated mTORC1 activity in OB mice, rapamycin treatment did not improve either glucose tolerance or insulin tolerance. These results indicate that the insulin resistance of OB mice is mediated, in part, by factors other than mTORC1.
- insulin resistance
- signal transduction
- skeletal muscle
type 2 diabetes mellitus is a major public health problem affecting ∼18 million Americans and resulting in annual total health care costs of approximately $120 million (18). Common to all type 2 diabetes patients is insulin resistance, a pathophysiology that precedes the onset of type 2 diabetes and is due to a reduction of insulin action in insulin-sensitive tissues. Because skeletal muscle is the primary site for postprandial glucose disposal, it is thought that this important metabolic tissue plays a major role in the development of insulin resistance and type 2 diabetes. Despite the important role skeletal muscle plays in glucose homeostasis, our knowledge of insulin signal transduction in skeletal muscle has not advanced as rapidly as with culture cells.
Several studies using cultured cells have shown that the activation of the mammalian target of rapamycin (mTOR) signaling pathway contributes to the development of insulin resistance (3, 29, 33–35). mTOR is an insulin-responsive and amino acid-sensitive Ser/Thr kinase whose activity is essential for cellular growth (4, 38). mTOR has been shown to contribute to insulin resistance through a negative feedback mechanism by promoting the Ser phosphorylation and subsequent degradation of insulin receptor substrate-1 (IRS-1) (5, 29, 35, 39). Evidence indicating that mTOR mediates the development of insulin resistance is strengthened by the ability of rapamycin to prevent Ser phosphorylation of IRS-1 (3, 29, 35, 39), as well as enhance PKB activity (33). Whether or not the activation of mTOR produces the insulin resistance associated with obesity in intact animals and humans has not been established. Furthermore, it is not known whether inhibiting mTOR activity with rapamycin, a highly specific inhibitor of mTOR, can improve insulin sensitivity.
Our knowledge of mTOR signaling has been greatly enhanced in the past few years by the discovery of numerous mTOR binding proteins. It is now well established that mTOR exists in at least two complexes, mTOR complex 1 (mTORC1) and mTORC2. mTORC1 contains raptor, and its activity is inhibited by rapamycin (15), whereas mTORC2 is rapamycin insensitive and contains rictor, the kinase responsible for phosphorylating PKB on Ser473 (26). mTORC1 also contains PRAS40 (2, 16), a novel target of PKB whose phosphorylation enhances mTORC1 activity (25, 37). In the basal state, PRAS40 binds tightly to mTORC1, repressing its kinase activity. Upon stimulation with insulin or growth factors, PRAS40 is phosphorylated by PKB and released from mTORC1, a process that enhances mTORC1 activity (25, 37). PKB can also enhance mTOR activity by promoting the phosphorylation of tuberous sclerosis complex 1 and 2 (TSC1/2). Phosphorylation of TSC1/2 enhances the GTP loading of Rheb (ras homolog enriched in brain), a process that appears to be necessary for increasing the kinase activity of mTORC1 (38). Because mTORC1 has been implicated in the development of insulin resistance and PKB regulates mTORC1 activity, it is possible that basal PKB activity is increased with insulin resistance.
The purpose of the present study was to determine whether elevated basal PKB/mTORC1 signaling is associated with skeletal muscle insulin resistance and whether or not rapamycin treatment can reverse in vivo insulin resistance. Although numerous cell culture studies have shown that activating mTORC1 impairs insulin signaling, the evidence implicating a role of mTORC1 in whole body insulin resistance is limited. Our hypothesis is that basal PKB/mTORC1 signaling will be elevated in muscles of ob/ob (OB) mice, an established model for obesity-related insulin resistance. We also hypothesize treating OB mice with rapamycin will improve in vivo insulin action.
All animal care and surgery were in accordance with the National Institutes of Health's Guide for Care and Use of Laboratory Animals [DHEW (NIH) Publication No. 85-23]. All experimental protocols were approved by the University Committee for the Use and Care of Animals.
Surgical procedures for OB and wild-type mice.
Male OB mice (n = 16) and age-matched male wild-type (WT) mice (n = 24) (Jackson Laboratories, Bar Harbor, ME) were caged individually, maintained on a 12:12-h light-dark cycle, and fed normal chow and water ad libitum. To harvest skeletal muscle, mice were anesthetized with a 1:1:1 mixture of promace, ketamine hydrochloride, and xyalzine by an intraperitoneal injection (0.015 ml/10 g body wt). Ten minutes before administration of the anesthetics, mice were injected with either insulin (0.25 U/kg body wt) or an equivalent volume of saline (0.9% NaCl). At 15 min following insulin or saline injections, quadriceps muscles were rapidly dissected, frozen in liquid nitrogen, and stored at −80°C until analysis.
Preparation of muscle extracts.
The frozen quadriceps muscles from all mice were manually ground with a porcelain mortar and pestle chilled in liquid N2. Powdered muscles were homogenized on ice using a motor-driven tissue grinder (Teflon-glass) in Cell Extraction Buffer (Biosource) (10 ml:1 g, buffer volume: muscle weight ratio) containing 0.1 mM dithiothreitol, 500 nM microcystin-LR, 0.1 mM phenylmethylsulfonyl fluoride, and 10 μg/ml each of aprotinin, leupeptin, and pepstatin-A. The homogenates were rotated at 4°C for 1 h and then centrifuged at 9,000 g for 30 min at 4°C. The protein concentrations of the supernatants were determined by the BCA method (Pierce, Rockford, IL). The remaining skeletal muscle extract was utilized for electrophoretic analysis and immunoblotting experiments.
Electrophoretic analyses and immunoblotting.
Skeletal muscle extracts and molecular weight standards (Bio-Rad, Hercules, CA) were subjected to SDS-PAGE. The proteins were then electrophoretically transferred to Immobilon membranes and immunoblotted with phospho-specific antibodies to S6K, mTOR, PKB, and PRAS40. The pThr389 S6K antibody recognizes S6K when phosphorylated on Thr389, a site that is phosphorylated by mTOR (23) and, therefore, an ideal measure of mTORC1 kinase activity (7). The pSer2448 mTOR antibody recognizes mTOR only when phosphorylated at Ser2448, a site whose phosphorylation is associated with increased mTORC1 kinase activity (19, 27, 28). The pSer473 PKB and pThr308 PKB antibodies recognize PKB when phosphorylated on Ser473 and Thr308, respectively, two sites whose phosphorylation is necessary for full PKB activity (1). And the pThr246 PRAS40 antibody recognizes PRAS40 when phosphorylated at Thr246. Phosphorylation of PRAS40 leads to increased mTORC1 kinase activity (25, 37). To account for changes in the total amounts of the signaling proteins, the phospho-specific immunoblots were stripped and reprobed with antibodies that recognize total S6K, mTOR, PKB-β, and PRAS40, respectively. After washing the membranes, we detected the light generated by the alkaline-phosphatase-conjugated secondary antibody and the CDP-Star reagent (Western Lightning; Perkin Elmer, Waltham, MA) using X-ray film (Kodak XAR-5). Relative signal intensities of the bands for pThr389 S6K, pSer2448 mTOR, pSer473 PKB, pThr308 PKB, and pThr246 PRAS40 were determined by using a digital densitometer (Fotodyne, Hartland, WI). All antibodies were from Cell Signaling Technology (Beverly, MA) unless noted otherwise.
Glucose and insulin tolerance testing.
To assess the role of mTORC1 on in vivo insulin sensitivity, all mice were subjected to two glucose tolerance (GT) and insulin tolerance (IT) tests. Two hours prior to each test, mice received an intraperitoneal injection of rapamycin (0.005 mg/g body wt) or an equivalent volume of sterile 0.9% saline. GT was assessed by injecting mice with glucose (1 mg/g body wt) and measuring glucose from blood collected via the tail vein at 0, 15, 30, 45, 60, and 90 min following the glucose injection. IT was assessed by injecting mice with insulin (0.25 U/kg body wt) and measuring glucose from blood collected via the tail vein at 0, 15, 30, 45, and 60 min following the glucose injection. The area under the curve (AUC) for GT and IT was calculated to assess the correlation between mTORC1 activity with GT and IT, respectively.
All data were analyzed using ANOVA. To detect statistical significance for basal and insulin-stimulated mTORC1 signaling, a 2 × 2 ANOVA (genotype × insulin) was used. To detect statistical differences for the effect of rapamycin on GT and IT, a 2 × 2 ANOVA (genotype × rapamycin) with repeated measures for time of glucose measure (0, 15, 30, 45, 60, or 90 min) was used. Following a significant F ratio, a priori mean comparisons were conducted using Fisher's least significant difference post hoc test. To detect statistical significance for AUC for GT and IT, one-way ANOVA was used. Data are expressed as means ± SE, and the level of statistical significance was set at P ≤ 0.05.
Phosphorylation of S6K on Thr389.
Phosphorylation on Thr389 of S6K is mediated by mTORC1 and serves as an ideal measure of mTORC1 (7, 23). Therefore, we have used the phosphorylation of S6K on Thr389 as a surrogate measure of mTORC1 kinase activity by using the phospho-specific antibody, pThr389 S6K. Fig. 1A shows a representative immunblot prepared with the pThr389 S6K phospho-specific antibody. As shown in Fig. 1B, basal phosphorylation of S6K on Thr389 is increased significantly in quadriceps muscle extracts prepared from OB mice compared with WT mice. Following insulin stimulation, pThr389 S6K immunoreactivty was significantly higher compared with the basal state; however, no significant differences were detected for pThr389 S6K immunoreactivity in muscle extracts of OB compared with WT mice (Fig. 1B). As shown in Fig. 1A, an immunoblot prepared with an antibody that recognizes all S6K regardless of phosphorylation demonstrates that changes in pThr389 S6K immunoreactivity are not due to differences in expression of the kinase, as total S6K immunoreactivity is not significantly different in the basal or insulin-stimulated state in muscle extracts prepared from OB compared with WT mice. These results provide evidence that basal mTORC1 activity is increased in skeletal muscles of OB mice.
Phosphorylation of mTOR on Ser2448.
Phosphorylation on Ser2448 of mTOR is highly associated with increased mTORC1 kinase activity (7, 19, 27, 28). Therefore, we have used the phosphorylation of this site as an additional measure of mTORC1 kinase activity by using the phospho-specific antibody, pSer2448 mTOR. Fig. 2A shows a representative immunblot prepared with the pSer2448 mTOR antibody. As shown in Fig. 2B,when normalized to the total amount of mTOR present in the samples, basal phosphorylation of mTOR on Ser2448 is increased significantly in quadriceps muscle extracts prepared from OB mice compared with WT mice (P ≤ 0.05). In WT mice, insulin promoted a significant increase in pSer2448 mTOR immunoreactivty compared with the basal state (P ≤ 0.05); however, in OB mice, no significant differences were detected for pSer2448 mTOR immunoreactivity when basal and insulin-stimulated samples were compared. As shown in Fig. 2A, an immunoblot prepared with an antibody that recognizes all mTOR regardless of phosphorylation demonstrates that total mTOR content in muscles from OB mice is similar to WT mice. In summary, these results examining mTOR directly provide further support that basal mTORC1 signaling is elevated in muscle of OB mice compared with WT mice.
Phosphorylation of PRAS40 on Thr246.
PRAS40 binds mTORC1, inhibiting its kinase activity (25, 37). Upon stimulation with insulin, PRAS40 is phosphorylated on Thr246 by PKB (2, 16) and released from mTORC1, a process that mediates an increase in mTORC1 kinase activity (25, 37). We have used the phosphorylation of PRAS40 on Thr246 as an additional measure of mTORC1 kinase activity by using the phospho-specific antibody, pThr246 PRAS40. In the basal state, pThr246 PRAS40 immunoreactivity is increased significantly in quadriceps muscle extracts prepared from OB mice compared with WT mice (Fig. 3B). Following insulin stimulation, pThr246 PRAS40 immunoreactivity is similar in muscle extracts of OB compared with WT mice (Fig. 3B). As shown in Fig. 3B, in both OB and WT mice, insulin produced a significant increase in pThr246 phosporylation (P ≤ 0.05). When the pThr246 PRAS40 immunoreactivity is normalized to total PRAS40 immunoreactivity and expressed as a ratio, the difference in basal PRAS40 phosphorylation between muscles of OB and WT mice persists, as does the potent effect of insulin. These results provide further support for our hypothesis that basal mTORC1 activity is increased in skeletal muscles of OB mice.
Phosphorylation of PKB on Thr308 and Ser473.
PKB promotes mTORC1 activity by phosphorylating PRAS40 (25, 37) and TSC1/2 (38). Because we demonstrate that basal mTORC1 activity (pThr389 S6K and pSer2448 mTOR immunoreactivity) and basal PRAS40 phophorylation were elevated in muscles of OB mice, we assessed PKB phosphorylation on Thr308 and Ser473, two sites whose phosphorylation are thought to be necessary for full PKB kinase activity (1), by immunoblotting with the phospho-specific antibodies pThr308 PKB and pSer473 PKB. A representative immunoblot prepared with pThr308 PKB antibody and the pSer473 antibody are shown in Figs. 4A and 5A, respectively. As shown in Figs. 4B and 5B, basal phosphorylation of PKB on Thr308 and Ser473 is increased significantly in quadriceps muscle extracts prepared from OB mice compared with WT mice. Following insulin stimulation, PKB phosphorylation is significantly increased in muscles of both OB and WT mice (Figs. 4B and 5B). As expected, insulin-stimulated phosphorylation of PKB on Thr308 and Ser473 are significantly lower in muscle extracts of OB compared with WT mice (Figs. 4B and 5B). As shown in Figs. 4A and 5A, immunoblots prepared with an antibody that recognizes all PKB-β regardless of phosphorylation demonstrates that changes in pThr308 PKB and pSer473 PKB immunoreactivities are not due to differences in expression of the kinase. These results indicate activated PKB in the basal state most likely mediated a portion of the increase in basal mTORC1 activity.
Because mTOR appears to produce insulin resistance in cultured cells by promoting the degradation of IRS-1 (5, 29, 35, 39), we assessed IRS-1 content in muscle from WT and OB mice using a highly sensitive ELISA. Similar to previous studies (6, 13), we demonstrate a 50% decline in IRS-1 levels in muscles from OB compared with WT (Fig. 6). These findings are consistent with the idea that overactive mTORC1 activity leads to lower IRS-1 content.
In vivo effects of rapamycin on glucose and insulin tolerance.
Little direct evidence supports a role for mTORC1 in the development of in vivo insulin resistance in intact animals or humans. Therefore, we examined the effect of acutely inhibiting mTORC1 with rapamycin 2 h before GT and IT tests in WT and OB mice. As expected, during the GT test, OB mice exhibited greater glucose values at baseline and at all time points after glucose administration compared with WT mice (Fig. 8). The higher glucose values during the GT test in OB mice resulted in greater total area under the GT curve compared with WT mice (⇓Table 2). The relationship between the area under the GT curve and mTORC1 activity (pThr389 S6K immunoreactivity) was very weak (R = 0.16). Although treating mice with rapamycin abolishes mTORC1 activity (Fig. 7), the inhibitor did not alter blood glucose values during the GT test (Fig. 8) or total area under the GT curve (Table 2) in either WT or OB mice. As expected, during the IT test, OB mice exhibited greater glucose values at baseline and at all time points after insulin administration compared with WT mice (Fig. 8). The higher glucose values during the IT test in OB mice resulted in greater total area under the IT curve compared with WT mice (Table 2). The relationship between the area under the IT curve and mTORC1 activity (pThr389 S6K immunoreactivity) was statistically significant (R = 0.44, P ≤ 0.05), suggesting that mTORC1 may play a role in the development of insulin resistance. Despite this moderately strong positive relationship, rapamycin did not alter blood glucose values during the IT test (Fig. 8) or total area under the IT curve (Table 2) in either WT or OB mice. Taken together, these results indicate that acutely inhibiting mTORC1 does not alter insulin sensitivity, despite a significant positive relationship between mTORC1 activity and area under the IT curve.
Numerous studies of cultured cells indicate that prolonged activation of the mTORC1 signaling pathway produces insulin resistance. However, the role that mTOR signaling plays in the development of in vivo insulin resistance in intact animals and humans is not established. Therefore, we have examined the activity of the mTORC1 signaling pathway in skeletal muscle of OB mice by assessing the phosphorylation of S6K on Thr389, a site directly phosphorylated by mTORC1. We have also assessed phosphorylation of mTOR on Ser2448, a site whose phosphorylation is associated with increased mTORC1 activity (7, 19, 27, 28). Furthermore, we have studied the effect of rapamycin treatment on in vivo insulin action in WT and OB mice. Our results clearly demonstrate that mTORC1 is constitutively activated in muscles of OB mice and associated with increased basal PKB activity (phosphorylation on Thr308 and Ser473). However, abolishing mTORC1 activity with rapamycin did not improve the insulin sensitivity in WT or OB mice.
The observation that an infusion of branched-chain amino acids decreased insulin-stimulated glucose disposal in human subjects (10, 22, 32) was the first indication that mTORC1 signaling mediates insulin resistance. Subsequent work by Patti et al. (20) established that amino acids reduced IRS-1 tyrosine phosphorylation and PI3K activity in L6 muscle cells and hepatocytes stimulated with insulin. More recent studies of cultured cells demonstrate that mTORC1 kinase activity mediates insulin resistance by reducing IRS-1 tyrosine phosphorylation, enhancing IRS-1 serine phosphorylation, and decreasing IRS-1 levels (11, 21, 34), a process that appears to be dependent on mTORC1 as raptor knockdowns prevent Ser phosphorylation of IRS-1 (35).
To the best of our knowledge, no studies have examined the role of mTORC1 in genetic models of obesity-related insulin resistance. In the present study, we demonstrate that mTORC1 signaling is elevated in the basal state in muscles from OB mice compared with WT mice. Similar to our results, S6K knockout mice are resistant to diet-induced insulin resistance (36), and rats fed a high-fat diet exhibited increased basal phosphorylation of mTOR on Ser2448 (14), a site phosphorylated by S6K (7). The present study expands upon studies of intact animals (14, 36) and cultured cells (3, 5, 35, 39), by demonstrating that mTORC1 activity is positively related to the degree of in vivo insulin resistance. Although our data are correlative, we provide further support for mTORC1 playing a role in the development of insulin resistance in a well-established model of obesity, the ob/ob mouse.
A novel finding of the present study is that basal PKB activity is increased in skeletal muscle of OB mice. The importance of this observation is underscored by the fact that PKB regulates mTORC1 activity (38). In muscles of OB mice, we demonstrate that the phosphorylation of PKB on Ser308 and Thr473 is increased compared with WT mice. This increase in basal PKB activity is most likely due to the well-documented hyperinsulinemia observed in OB mice, despite a significant reduction in IRS-1 levels (see Fig. 6). PKB regulates mTORC1 activity by two proposed mechanisms. First, PKB has been suggested to enhance mTORC1 activity by phosphorylating TSC1/2, a process that results in greater GTP loading of the mTOR interacting protein Rheb. Rheb·GTP is essential for mTOR activity (38). Second, PKB has been suggested to enhance mTORC1 activity by phosphorylating PRAS40, a process that releases PRAS40 from mTORC1, resulting in increased mTORC1 kinase activity (25). Another novel finding of the present study is that basal PRAS40 phosphorylation on Thr246 is increased in muscles of OB compared with WT mice, a process mediated by PKB that likely contributes to increased mTORC1 kinase activity. mTORC1 can also be activated independent of PKB by branched-chain amino acids, particularly leucine (38). She et al. (30) have recently demonstrated that OB mice have elevated plasma branched-chain amino acid levels, which would result in increased mTORC1 activity. Taken together, our findings and those of She et al. (30) indicate that skeletal muscle mTORC1 activity is increased in obesity-related insulin resistance by PKB-dependent and PKB-independent mechanisms.
Because we demonstrate that basal mTORC1 activity is elevated in muscles from OB mice, it is a reasonable hypothesis that inhibiting mTORC1 signaling would improve in vivo insulin action. This point is underscored since data from cell culture studies demonstrate that rapamycin reverses insulin resistance (3, 5, 35, 39). We have also demonstrated that incubating rat epitrochlearis muscles with amino acids impairs insulin-stimulated PKB activity, a process reversed by rapamycin treatment (24). Despite these in vitro findings, treating WT and OB mice with rapamycin prior to assessing GT and IT did not improve in vivo insulin sensitivity (see Fig. 8) despite abolishing mTORC1 activity (see Fig. 7). Although the discrepancy between the in vitro (cultured cells, muscle incubations) and in vivo effects of rapamycin is difficult to explain, it underscores the importance of studying whole body insulin action rather than cultured cells or incubated intact muscles. One potential explanation for the lack of an in vivo effect of rapamycin may be the length of time mTORC1 was activated. Prolonged mTORC1 activation leads to insulin resistance by reducing IRS-1 levels via Ser phosporylation (5, 29, 35, 39). Although acute rapamycin treatment abolishes mTORC1 activity, it would not be expected to restore the 50% reduction in IRS-1 levels we observed in muscles from our OB mice. The magnitude of the present decline in IRS-1 is similar to previous studies of OB mice (6, 13) and obese humans (9, 17). Perhaps chronic treatment with rapamycin might have restored IRS-1 levels, but these experiments are problematic as mTOR plays a key role in the mRNA translation initiation of proteins necessary for insulin action. In fact, chronic in vivo rapamycin treatment can impair insulin action (8). The lack of an acute rapamycin effect on insulin sensitivity in our OB mice suggests that kinases other than mTOR disrupt insulin signaling. Herschkovitz et al. (12) demonstrate that IRS-1 contains at least seven Ser phosphorylation sites and that PKCθ, p38 MAPK, and PKCζ all converge on IKKβ, a Ser kinase that phosphorylates IRS-1. Rapamycin treatment would be expected to prevent Ser phosphorylation of IRS-1 on Ser636 (35), an established mTORC1 phosphorylation site, but altering a single specific Ser phosphorylation site of IRS-1 is not sufficient to disrupt IRS-1 signaling (11). However, Sugita et al. (31) clearly demonstrates inducible nitric oxide synthase (iNOS) reduces IRS-1 levels in skeletal muscle and treating ob/ob mice with the iNOS inhibitor N6-(1-Iminoethyl)-l-lysine dihydrochloride restores IRS-1 to levels found in muscles of WT mice. Finally, we cannot rule out the possibility that the lack of an in vivo effect of rapamycin on GT and IT may be due to nonspecific effects of the inhibitor that may not be present in in vitro cell culture studies.
Perspectives and Significance
In conclusion, we demonstrate that basal PKB/PRAS40/mTORC1 signaling is elevated in skeletal muscle of OB mice compared with WT mice. We also demonstrate a strong relationship between mTORC1 activity in skeletal muscle and the degree of whole body insulin resistance. However, treating OB and WT mice with rapamycin, a highly specific inhibitor of mTORC1, does not improve in vivo insulin action, indicating that other factors in addition to mTORC1 mediate insulin resistance. These findings highlight the importance of examining whole body insulin action to confirm cell culture studies, identifying novel factors regulating insulin signaling. Future studies examining the role of mTORC1 signaling and the effects of rapamycin treatment in other models of obesity-related insulin resistance are warranted.
This work was supported by 7 R15 DK065645-02 from the National Institutes of Health.
The authors are grateful to Thurl Harris for his help in preparing this manuscript.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2008 the American Physiological Society